Jurnal Geologi , Vol. 4 No. 4 Desember 2009: 229-238

Emergence of from the Crater Lake of Kelud Volcano,

Sr i Hi d a y a t i , aH m a d Ba S u k i , kr i S t i a n t o , and iy a n mu l y a n a

Centre for Volcanology and Geological Hazard Mitigation, Geological Agency Jln. Diponegoro 57 40122

Ab s t r A c t Kelud Volcano (+1731 m) in East Java is one of the most active and dangerous volcanoes in Indonesia. A large lake occupies the summit crater. Historical eruptions generally only lasted for a very short time, mostly no longer than a few hours. The outburst is usually accompanied by pyroclastic flows. On August 2007, the activity of the volcano was initiated by the increase of the temperature of lake water and the change of the colour from typical green to yellow. Activities of the volcano are discussed following the swarms of volcano-tectonic (VT) earthquakes on September 10th, September 26th to 29th, and October 24th to November 2nd. On September 26th to 29th, hypocentral distribution of those VT shifted from 5 km deep to just beneath the crater. The highest number of VT earthquakes occurred on November 1st attaining 50 events, then followed by a swarm of B-type events, where the number reached 1437 events in a day. The volcanic activity peaked on November 3rd when seismic records became saturated, which then was preceded by a sharp increase of lake temperature and a sudden deflation of radial tilt. It suggests thatthe lava extrusion forming a lava dome was taking place. Keywords: Kelud Volcano, crater lake, seismic activity, VT earthquakes, lava dome

Sa r i Gunung Kelud di Jawa Timur merupakan salah satu gunung api berdanau kawah di puncaknya yang sangat aktif dan berbahaya di Indonesia. Sejarah letusannya tercatat sebagai letusan-letusan yang berlangsung sa- ngat singkat, hanya beberapa jam. Pada bulan Agustus 2007 aktivitas Gunung Kelud mengalami peningkatan dengan teramatinya perubahan warna danau kawah dari kehijauan menjadi kekuningan. Pemunculan gempa volcano-tectonic (VT), pada 10 September, 26 - 29 September, dan 24 Oktober - 2 November akan dibahas dalam makalah ini. Sebaran hiposentrum gempa VT pada 26 - 29 September bergerak naik dari kedalaman 5 km menuju tepat di bawah kawah. Jumlah gempa VT tertinggi terjadi pada 1 November mencapai 50 kejadian, kemudian diikuti oleh terjadinya gempa tipe B dengan jumlah 1437 kejadian. Aktivitas Gunung Kelud mencapai puncaknya pada tanggal 3 November ketika rekaman seismik tersaturasi, yang didahului oleh kenaikan suhu air kawah secara tajam dan deflasi yang tiba-tiba pada komponen tilt sumbu radial. Saat itu diperkirakan mulai berlangsung ekstrusi lava yang membentuk kubah lava. Kata kunci: Gunung Api Kelud, danau kawah, aktivitas seismik, gempa VT, kubah lava

IntroductIon tions are in 1586, 1901, 1919, 1951, 1966, and 1990. Those eruptions were characterized by initial Geographically the Kelud Volcano is located at phreatomagmatic eruptions followed by explosive 7o56’S and 112o18’E. The volcano lies within two ones which produced pyroclastic flows, ash-fall, regencies, i.e. and , in East Java (Figure and lapilli. Furthermore, the explosive eruptions of 1). The summit elevation of this is the volcano usually take place in a short time. Lava 1731 m asl. and a large lake occupies the summit dome formations were also recorded in 1376 and crater. 1920 (Kusumadinata, 1979). Historical eruption of the Kelud Volcano was The Centre for Volcanology and Geological Haz- recorded firstly in the year of 1000. The big erup- ard Mitigation (CVGHM) has monitored the Kelud

Naskah diterima: 21 Oktober 2008, revisi kesatu: 06 Maret 2009, revisi kedua: 10 Agustus 2009, revisi terakhir: 09 Oktober 2009 229 230 Jurnal Geologi Indonesia, Vol. 4 No. 4 Desember 2009: 229-238

Java Sea Minakami classification. Minakami (1974) classi- fied volcanic earthquakes according to the location of their foci and the nature of earthquakes motion into: A-type, B-type, explosion earthquakes, and East Java tremors. The A-type earthquakes originate beneath the volcanoes at the depth ranging from 1 km to 20 G. Arjuno Welirang G. Lawu km. Their waveform cannot be distinguished from G. Kelud G. Bromo G. Iyang Argopuro G. Ijen shallow tectonic earthquakes. The P- and S-phases G. Lamongan G. G. Raung are also clearly defined. The B-type earthquakes N occur at shallow depths less than 1 km beneath the 0 50 100 km Indian Ocean active crater and their S-phases are not clear. Fur- thermore, explosion earthquakes always accompany Figure 1. Location map of the Kelud Volcano. an individual explosive eruption, while volcanic tremors have a form of an irregular sinusoid of Volcano since 1925 using one station of a single rather long duration. seismometer. In 1987, the number of seismometer Meanwhile, on the basis of source mechanism, was extended into three. All of them were destroyed Latter (1981) proposed the earthquakes originating during the eruption of February 1990. There were at Ruapehu and Ngauruhoe Volcanoes into two some replacements of the seismometers after the groups: tectonic and volcanic earthquakes. The vol- eruption. In 2006, a tiltmeter was installed on the canic earthquakes take place in a heat-weakened or west flank and in early 2007 another tiltmeter was partially molten material by some extended source added at the southwest flank. Since April 2007, five mechanisms. Generally, their onset is emergent seismometers (L4-C, 1 Hz) and two tiltmeter stations with poorly-defined phases. This category covers were set up to monitor the volcano. The seismic Minakami of B-types, explosion earthquakes, and stations are KWH, KLD, LRG, SMG and UMB. tremor. Tectonic type is characterized by sharp on- The data are transmitted to Kelud Observatory in set and well-defined phases. In order to distinguish Margomulya Village at a distance of about 8 km from between earthquakes of tectonic type which origi- the active crater using a telemetry system. nate on or beneath a volcano and those that occur Since 2006, a sensor of temperature has been at some distances, a term of volcano-tectonic (VT) installed at the crater lake to monitor the surface of earthquakes is proposed. This category includes A- lake water at 10 m and 15 m deep. Moreover, a pe- type of Minakami’s. VT earthquakes take place in riodic monitoring of chemical contents of the crater a competent rock as a result of some instantaneous lake water has a significant support to the activity of source mechanism as the tectonic one do. In this pa- the volcano. In August 2007, the volcanic activity per, the term of VT earthquakes will be used instead increased as shown by the change of the water colour of A-type, since this term is recently used widely. from typical green to yellowish. This change was ac- During the normal state of activity, the occurrence of VT earthquake is usually less than 5 events/month, companied by the increase of the CO2 concentration and the temperature of lake water. meanwhile the B-type and other type of volcanic The paper will discuss about the seismic activ- earthquakes are very rarely recorded (Figure 2). ity, hypocentral distribution, ground deformation, As shown in Figure 2, on September 11th, 2007 and other significant precursors that precede the the number of VT earthquake reached 13 events emergence of the lava dome. within 5 hours. Using GrHypo program (Nishi, 2005) with onset of 4 P-waves and 1 S-wave, their hypocentres were determined. The result showed se I s m I c A c t I v I t y o f t h e Ke l u d v o l c A n o the distribution ranges at depth of about 2 to 4 km p r e c e d I n g t h e e m e r g e n c e o f l A v A d o m e below the crater (Figure 3.a). The occurrence of VT earthquakes continued CVGHM is used to distinguish volcanic earth- in average one event a day. During September 26th quakes observed at active volcanoes based on the to 29th, the number of VT earthquakes increased Emergence of Lava Dome from the Crater Lake of Kelud Volcano, East Java (S. Hidayati et al.) 231

Level III Level IV Level III 400 B-type Level II Level I 200

0 40 7/1 7/15 7/29 8/12 8/26 9/9 9/23 10/7 10/21 11/4 11/18 VT 20

0 4 7/1 7/15 7/29 8/12 8/26 9/9 9/23 10/7 10/21 11/4 11/18 local tectonics 2 umberof events N 0 20 7/1 7/15 7/29 8/12 8/26 9/9 9/23 10/7 10/21 11/4 11/18 tectonics 10

0 7/1 7/15 7/29 8/12 8/26 9/9 9/23 10/7 10/21 11/4 11/18 2007 Date

Figure 2. The number of VT and B-type earthquakes recorded at the Kelud Volcano during July – November 2007. The vertical lines and horizontal arrows represent the change of level activity (level II, III, and IV) as declared by CVGHM. remarkable to 61 events. However, no significant swarm was followed by the B-type. The number is change was indicated on their hypocentral dis- 89 events. On November 1st, the VT earthquakes tribution (Figure 3.b). After September 29th, the reached 55 events while the B-type reached re- occurrence of VT earthquakes still continued but markably to 1437. Meanwhile, tremor attained 274 decreasing in number. Their foci shifted to the depth events with dominant amplitude about 0.2 to 5 mm. of about 0.5 from 1 km beneath the crater lake. It was On November 2nd, number of the VT earthquakes supposed that a migration of magma to the surface reached 52 events while the B-type 690 events. occurred (Figure 3.c). The hypocentral distributions of VT earthquakes On October 16th, swarm of 306 of B-type events of October 31st, November 1st and 2nd are about 0.5 occurred within 7 hours. Later, number of B-types to 5.0 km of depth (Figure 3.d). earthquakes sharply decreased, until October 21st Continues tremor was recorded since Novem- there were only 4 events recorded. After 2 days in ber 2nd at 11:07 WIB (local time) and followed by quiescence, on October 24th, the VT earthquakes oc- the saturated seismic event on November 3rd, at curred about 10 events. The occurrence was directhly 15:55 for about 10 minutes. In the next morning, followed by the B-type earthquakes. in a clear weather, a fuming small black dot raised. On October 31st, other swarms of VT earth- It was observed in the middle of crater lake. The quakes with large amplitude attained 31 events, phenomena was associated with the emergence of and their (S-P) time were less than 1 second. This a lava dome. 232 Jurnal Geologi Indonesia, Vol. 4 No. 4 Desember 2009: 229-238 -November

th West - East (km) West West - East (km) West

-2 0 2 , and (d) October 24 October (d) and , th -3 -2 -1 0 1 2 3

-15 1 0 3

2

-1 -3 -2 st

0 2 -2 -4 South - North (km) North - South Depth (km) Depth (d)

; (c) October 1 October (c) ; th

-29 th

West - East (km) West West - East (km) West -2 0 2

; (b) September 26 September (b) ; th

-3 -2 -1 0 1 2 3

1 0 3 0 2 2 -1 -3

-2 -2

-4 South - North (km) North - South Depth (km) Depth (c)

West - East (km) West West - East (km) West -2 0 2

-3 -2 -1 0 1 2 3

1 0

3 2 0 2 -1

-3 -2 -2 -4 South - North (km) North - South Depth (km) Depth (b) West - East (km) West West - East (km) West -2 0 2

-3 -2 -1 0 1 2 3

1 0 3

2 0 -1 2 -3 -2

-2 -4 Depth (km) Depth South - North (km) North - South (a) . Solid triangles and circles represent seismic stations (LRG, KLD, KWH, and SBG) and hypocentres of the VT earthquakes, respectively. VT . Solid triangles and circles represent seismic stations (LRG, KLD, KWH, SBG) hypocentres of the st 1 Figure 3. Hypocentral distribution of the VT earthquakes during activity on (a) September 10 September (a) on activity during earthquakes VT the of distribution Hypocentral 3. Figure Emergence of Lava Dome from the Crater Lake of Kelud Volcano, East Java (S. Hidayati et al.) 233

dI s c u s s I o n in August and September from the typical green to yellow and bluish. The lake temperature measure- Magma Supply System ments showed a gradual increase. The eruption of Kelud Volcano in 1990 was The VT earthquakes often precede a swarm of preceded by a sharp increase in number of VT earth- the B-type, then followed by repeated explosive quakes several days before the outbreak (Lesage and eruptions similar to the Sakurajima Volcano (Kamo, Surono, 1995). However, the significant increased in 1978). Meanwhile, the swarm of B-type is inferred seismicity and lake temperature was detected three to be related to extrusion of lava to the summit crater months before the eruption. (Nishi, 1984). As explained by Latter (1981), the VT The precursor of the 1990 eruption was also earthquakes usually occur in a competent rock as a observed in 2007, i.e. the increase of seismicity result of some instantaneous source mechanisms. and the lake temperature. As mentioned above that The VT earthquakes of the Sakurajima Volcano emergence of lava dome in the middle of crater lake mostly originated outside the crater rim, enclosing was preceded by the increase of seismicity starting other type of volcanic earthquakes such as explo- in September 2007. Swarm of the B-type event on sion and the B-type earthquakes (Ishihara, 1990; October 16th (Figure 4) was preceded by an increas- Iguchi, 1994; Hidayati et al., 2007). Meanwhile, ing number of VT earthquakes on September 10th, the B-type earthquakes originate in partial molten 2007 (Figure 2). These occurrences were repeated material by some extended source mechanism. Fur- on October 31st when a number of VT increased thermore, in the Merapi Volcano, the occurrence of then followed by a swarm of the B-type event on VT earthquakes precedes the increase in number of November 1st. The number of the B-type exceeds MP events which are related to lava dome formation about 1437 events during the day (Figure 5). (Ratdomopurbo, 1995; Hidayati et al., 1998). On the other hand, increasing of the lake tem- It all explains that the occurrence of VT earth- perature was detected earlier. Since July 2007, an in- quakes at the Kelud Volcano may indicate there was tense degassing of the lake floor has been observed. a magma supply from deeper part to the depth about 3 It causes colour of the lake waters changed rapidly - 4 km beneath the crater lake (Figure 3). Meanwhile,

From: 2007/10/16 12:00:00 PM To: 11:59:59 PM KL4-UD Interval 12 hours Scale: 25 % of 40 (Max Amp.) bit:21

12:00 :20 :40 13:00 :20 :40 14:00 :20 :40 15:00 :20 :40 16:00 :20 :40 17:00 :20 :40 18:00 :20 :40 19:00 :20 :40 20:00 :20 :40 21:00 :20 :40 22:00 :20 :40 23:00 :20 :40

Figure 4. A swarm of the B-type took place on October 16th, 2007 recorded by Kawah station, located at about 400 m from the crater lake. 234 Jurnal Geologi Indonesia, Vol. 4 No. 4 Desember 2009: 229-238

From: 2007/10/31 12:00:00 PM To: 11:59:59 PM KL4-UD Interval 12 hours Scale: 25 % of 40 (Max Amp.) bit:21

12:00 :20 :40 13:00 :20 :40 14:00 :20 :40 15:00 :20 :40 16:00 :20 :40 17:00 :20 :40 18:00 :20 :40 19:00 :20 :40 20:00 :20 :40 21:00 :20 :40 22:00 :20 :40 23:00 :20 :40

From: 2007/11/01 12:00:00 AM To: 12:00:00 PM KL4-UD Interval 12 hours Scale: 25 % of 40 (Max Amp.) bit:21

00:00 :20 :40 01:00 :20 :40 02:00 :20 :40 03:00 :20 :40 04:00 :20 :40 05:00 :20 :40 06:00 :20 :40 07:00 :20 :40 08:00 :20 :40 09:00 :20 :40 10:00 :20 :40 11:00 :20 :40

Figure 5. The swarm of VT occurred on October 31st, 2007 followed by a swarm of the B-type, recorded by Kawah station which is located at about 400 m from the crater lake. the swarm of B-type events may show a migration was due to an ascending of contribution from the of magma through the conduit to the shallow part of deep hydrothermal system (Kunrat, 2009). volcano edifice. Moreover, chemical properties of lake water revealed two hydrothermal systems: the Emergence of Lava Dome deep and shallower ones. The increasing temperature The occurrence of VT earthquakes ceased in the and acidity of the lake water before 2007 eruption morning of November 1st (Figures 2 & 5). The ces- Emergence of Lava Dome from the Crater Lake of Kelud Volcano, East Java (S. Hidayati et al.) 235 sation of those VT was replaced by a swarm of the saturation lasted in 10 minutes. The sharp increase B-type earthquakes which inferred to be associated of temperature, the sudden deflation of radial tilt, and with a magma migration to the shallow part. The the saturated seismic event simply occurred almost swarm ended in early morning of November 2nd. at the same time. It seemed to indicate an extrusion On November 2nd, after the quiescence of seis- of lava was taking place. However, the extrusion micity for about 8 hours, tremor began to occur. could not be observed directly because of heavy rain. The amplitude of tremor became larger and then at The following day, on November 4th, was observed 14:43 it changed to a spasmodic tremor which lasted a small fuming black dot in the crater lake from for 70 minutes. The tremor was interspersed by the CCTV (close circuit television) camera (Figure 8; emission earthquakes, and both amplitudes gradually left). About two months of the high seismicity, which became larger. Meanwhile, at the same day started was much larger than the previous eruption in 1990, at 06:00 the lake temperature significantly increased it was expected that this condition would produce a (Figure 6) and then at surface and at the 10 m of bigger explosive one. In fact, it was a lava doming depth have the same value. in the middle of crater lake of the Kelud Volcano. On November 3rd, at about 12:00, the sharp Furthermore, the sensor of lake temperature was increased of lake temperature was noted (Figure broken when the lava dome emerged due to the heat 6). Meanwhile, at 12:13, the ground deformation exceeded its capability. measured by tiltmeter which is located about 600 m After November 3rd, the activity of the volcano away to the southwest from the active crater, showed gradually decreased. The VT and B-type earth- a sharp deflation of radial tilt (Figure 7). Moreover, quakes, the events supported a magma supply, has at 15:55, the seismic record became saturated. The ceased. Tremor continued to occur and intersperse

80

surface temp. 10 m depth 15 m depth

70 ) C o

( 60

50 Temperature

40

31/11/2007 06:0031/11/2007 18:0001/11/2007 06:0001/11/2007 18:0002/11/2007 06:0002/11/2007 18:0003/11/2007 06:0003/11/2007 18:0004/11/2007 06:0004/11/2007 18:00

Figure 6. The sharp increase of lake water temperature on November 3rd, 2007, before the lava dome emerged from the bot- tom of crater lake. Red triangle, green square, and orange circle represent the temperatures at the crater lake surface, in the depth of 10 m and 15 m, respectively. 236 Jurnal Geologi Indonesia, Vol. 4 No. 4 Desember 2009: 229-238

110 100 90 80 70 60 50 40

TANGENTIAL (mrad) 30

25 Lava extrusion began 20 15 10 Temperature (oC)

-100 phreatic eruption -104 on Nov 11, 2007 -108 -112 -116 phreatic eruption -120 on Nov 14, 2007 -124 -128 -132

RADIAL (mrad) -136 -140 -- 2007-09-29 2:53 2007-10-06 1:32 2007-10-13 0:12 2007-11-09 9:51 11/12/2007 21:11 2007-09-25 15:33 2007-10-02 14:13 2007-10-09 12:52 2007-10-16 11:32 2007-10-19 23:52 2007-10-23 11:11 2007-10-26 12:31 2007-10-29 23:51 2007-11-02 11:11 2007-11-05 22:31 2007-11-15 23:31

Figure 7. Tilt measurement at the distance about 600 m from the crater on the SW flank. A sharp deflation was observed at 12:13 (local time) on November 3rd, 2007.

Figure 8. The emergence process of lava dome from the crater lake of Kelud Volcano (taken from CCTV camera; courtesy of Local Authority of Kediri). with emission earthquakes. The seismic record be- Field observations were carried out on 6th, 9th, came saturated again on November 11th, 15:51 for 11th, and 17th of November 2007 and found out 36 minutes and on November 14th, 2007 at 03:47, that the dome grew very rapidly. The height of the for 17 minutes. At the same time, ash reached about dome was gradually raised. Every visual observa- 2500 m high above the volcano and some materials tions at the crater estimated that the height of the were ejected from the crater as recorded by CCTV dome was 70 m, 100 m, 120 m, and 150 m from camera. crater bottom. Emergence of Lava Dome from the Crater Lake of Kelud Volcano, East Java (S. Hidayati et al.) 237

After the 11th of November, the seismic record As mentioned in previous chapter that eruption was dominated by emission earthquakes. As shown of the Kelud Volcano was characterized by mostly in Figure 9, the daily number of the emission earth- explosive eruption and very rare lava dome forma- quakes showed a peak and took place at the end of tion. However, a question arises why after several November until the third week of December 2007. months of high seismicity on July to October 2007, Afterward, the number of emission earthquakes only formed a lava dome instead of explosive erup- gradually decayed. tion? From the seismicity point of view, it is possibly The observations at the crater repeated on April that there is no enough released-energy to break 22nd and August 26th, 2008 found out that the dome the magma barrier and explode. The other possibil- occupied almost the whole crater lake with 489 m ity, there was not enough magma supply from the in diameter and its height reached of about 216 m main magma chamber, although there was detected above the crater bottom. By these phenomena it magma migrated from about 4 km beneath the crater could be suggested that the growth of lava dome lake to the surface. seems to be ceased (Figure 9). The number of emis- Actually, a sharply increasing of flux of CO2 sion earthquakes also sharply decreased; it was less about 550 ton/day was detected on early September, than 5 events per day (Figure 10). 2007, where in a normal state it was only about 40 – 50 ton/day. Unfortunately, no further observation

250 on flux of CO2 was carried out afterward due to

April 22, 2008 Aug 26, 2008 safety reason. Furthermore, analysis of chemical properties of lake water (Kunrat, 2009) revealed 200 that degassed magma was originated from an old remained one, not a new one, since the magma only

150 produced heat and the pH of lake water remains about neutral as it is. Based on petrography analysis, Zaennudin and 100 Height of lava dome (m) Budianto (2009) figured out that the lava had mostly solidified before appearing to surface. The lava dome

50 of 2007 is composed of euhedral crystals of plagio- Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep 2007 2008 clase and pyroxene. It means that the crystallization Date process was on the state when no temperature and Figure 9. The growth of lava dome based on visual observa- pressure significant changed. Probably, this lava tion. dome is a continuation of the process eruption which occurred in 1990. In addition, at the end of 1990 eruption, a black clod was seen approximately 3 m 1800 in the middle of crater (Pratomo, Personal Comm., 1600 2008). It could be a lava plug at top of the conduit. 1400

1200

1000 co n c l u s I o n

800

600 In early August 2007, the activity of the volcano Number ofevents

400 began to increase. Swarms of VT earthquakes have preceded the emergence of lava dome. A magma 200 migration from deeper part to the shallow one was 0 Nov Dec Jan Feb Mar Apr May Jun Jul Aug detected from September to October 2007 by the 2007 2008 Date development of hypocentral distribution of VT. The Figure 10. The daily number of emission earthquakes . The sharp increase of temperature, the sudden deflation number peaked from the end of November until the third of radial tilt, and the saturated of seismic event week of December 2007 and then decreased. showed that there is a mass of magma coming up 238 Jurnal Geologi Indonesia, Vol. 4 No. 4 Desember 2009: 229-238 to the surface on November 3rd. It indicates that an Kamo, K., 1978. Some phenomena before the summit extrusion of magma forming lava dome is began. eruptions at Sakurajima volcano. Bulletin Volcanological Society of Japan, 23, p.53-64. (In Japanese with English Acknowledgments---The authors are grateful to Budi Pri- abstract) yanto, Ananto Bagus Kuncoro, and Khairil Huda, the vol- Kunrat, S.L, 2009. Geochemical and Thermodynamic cano observers of Kelud Volcano observatory for their help- Modeling of Volcanic Fluids and Their Interpretation ful assistances. Many appreciation to Ahmad Zaennudin for for Volcano Monitoring. Université Libre de Bruxelles fruitful discussions. Thanks to CVGHM for providing the (unpublished paper). Kusumadinata, K., 1979. Data Dasar Gunungapi Indonesia. fund through the fiscal budget of the year 2007. Direktorat Vulkanologi, Bandung. Latter, J.H., 1981. Volcanic earthquakes and their relationship to eruptions at Ruapehu and Ngauruhoe volcanoes. re f e r e n c e s Journal of Volcanology and Geothermal Research, 9, p.293 - 309. Hidayati, S., Iguchi, M., Ishihara, K., Purbawinata, M.A., Lesage, Ph. and Surono, 1995. Seismic precursors of the Subandriyo, Sinulingga, I.K., and Suharna, 1998. A February 10, 1990 eruption of Kelut volcano, Java. preliminary result of quantitative evaluation on activity Journal of Volcanology and Geothermal Research, 65, of Merapi volcano. Proccedings On Symposium on Japan p. 135-146. - Indonesia IDNDR Project, p. 165 - 180. Minakami, T. 1974. Seismology and volcanoes in Japan. In: Hidayati, S., Ishihara., K., and Iguchi, M., 2007. Volcano- Civetta, L., Gasfarini, P., Luongo, G., and Rapolla, A. tectonic Earthquakes during the Stage of Magma (Eds.), Physical Volcanology, p.1 - 27. Elsevier. Accumulation at the Aira Caldera, Southern Kyushu, Nishi, K., 1984. Volcanic B-type earthquake swarm preceding Japan. Bulletin Volcanological Society of Japan, 52 (6), volcanic explosion. Annual Disaster Prevention Research p.289-309. Institute, Kyoto University, 27B-1, p.29-34. (In Japanese Iguchi. M., 1994. A vertical expansion source model for the with English abstract). mechanisms of earthquakes originated in the magma Nishi, K., 2005. Graphical Hypocentre Calcuation software conduit of an andesitic volcano: Sakurajima, Japan. (GrHypo). Bulletin Volcanological Society of Japan, 39, p.49-67. Ratdomopurbo, 1995. Etude sismoloique du Vulcan Merapi Ishihara, K., 1990. Pressure sources and induces ground et formation du dome de 1994. PhD thesis, L’universite deformation associated with explosive eruptions at Joseph Fourier-Grenoble I, France, Unpublished. andesitic volcano: Sakurajima volcano, Japan. In: Ryan, Zaennudin, A. and Budianto, A., 2009. Prakiraan Bahaya M.P. (Ed.), Magma Transport and Storage, John Willey Erupsi G. Kelud (unpublished report). & Sons, Chischester, p.335 - 356